My Top Simple.ai Posts from 2025

It's been quite a year for simple.ai.

When I started this newsletter about a year ago, I wasn't sure what would resonate. I don't have a machine learning PhD, wasn’t one of the authors of the “Attention Is All You Need” paper and I’ve never worked at one of the frontier AI research labs like OpenAI or Anthropic. I just figured I'd write about what I was learning, building, and thinking about -- and hope some of it would be useful to others.

Turns out, a lot of you found it helpful. Simple.ai has grown to over 2 million subscribers now, which is honestly pretty surreal (and exciting!).

I think what's worked is that I write like I think -- in simple primitives, analogies, and practical examples. I'm a mere mortal sharing some thoughts with (I think) other mere mortals. Unless of course, Andrej Karpathy somehow wandered into this — you never know. Anyways, I’m here to help the millions who are trying to keep up with this technology and use it to improve their careers and companies.

As we close out 2025, I wanted to share the 5 favorite posts from this year.

These are the ones that meant the most to me -- either because they changed how I think about something, or because of the conversations they sparked.

If you're new, these will give you a sense of what I cover. If you've been here all along, it's a good reminder of how much the AI landscape has changed in just 12 months.

— @dharmesh

#1: How to compete with AI (and win)

This came from my TEDx Boston talk, which I had exactly 12 nights to prepare for. This YouTube video has had over 250,000 views (which is great), but what is more heart-warming is that it just passed 20,000 likes. If you are one of those people, THANK YOU!

The core idea in the video: It's not you versus AI. It's you to the power of AI.

I broke down why LLMs are basically "smart autocomplete," AI's three critical limitations, why banning AI is like banning spell check, and how to actually get AI adoption in organizations.

This one means a lot to me because I think it reframes the whole AI conversation more productively.

The reality is more nuanced. AI will take your job and give you one that's much better -- one with less rote work, fewer mind-numbing repetitive tasks, and more focus on what actually matters.

We're not less valuable because we use AI tools. We're more valuable because these tools amplify our capabilities.

Read the full post here

#2: How AI is reshaping content creation

This one tackled a paradox I keep thinking about: content creation is simultaneously getting easier and harder.

With AI video tools like Google’s Veo 3 creating hyper-realistic videos in minutes, and AI image generators like Nano Banana now creating images that are indistinguishable from reality, content creation is becoming 100x easier to create.

Naturally, when content is 100x easier to create, we’ll see 100x more content -- content that you and your brands are competing with on the algorithm.

This is making it exponentially harder for brands to stand out and driving the demand for skilled content creators.

My core advice: Focus on creating content that AI fundamentally cannot generate -- content rooted in your human experiences, your unique insights, your specific failures and lessons learned.

I also talked about why LinkedIn is still one of the most underrated career hacks, but the strategy has shifted:

• Then: Keep your profile updated with what you’ve done

• Now: Post regularly with your actual thoughts and insights

The key is authenticity over commercialization. Share what you think, not just what you've done.

Read the post here.

#3: Why AEO is the next marketing gold rush

This one was about Answer Engine Optimization (AEO) -- optimizing your content to show up in AI-powered answer engines like ChatGPT, Gemini, Claude and others.

I'm including this because I think AEO is one of those shifts that will matter more than people realize right now. Even HubSpot is feeling it. We’ve even recently acquired XFunnel to boost our AEO efforts.

The data backing this decision has been pretty stark. It’s commonly known now that organic traffic from traditional SEO has declined 20-40% across industries over the past couple of years.

People's search behavior has changed. Instead of going to Google and clicking through 10 links, they're going directly to ChatGPT and getting answers immediately.

This creates what I call a "binary outcome" problem. Either you show up in the AI's answer, or you might as well not exist.

I shared six tactical ways to optimize: enable AI crawling, track your AI traffic, reframe your content structure to question-answer formats, apply the "Deserve to Rank" test, leverage classic SEO best practices, and invest in Reddit and social discussions.

We're still early in the AEO game. Just like the early days of SEO, the businesses that move quickly will have significant advantages.

Read the post here.

#4: Why I still code

This was a bit different from my usual posts -- more of a personal rant about a fundamental choice we all face: Do you spend your time fixing your weaknesses, or amplifying your strengths?

I get asked all the time: "Why do you still code? Couldn't you hire people to do that?"

My answer is simply because it's what I'm good at. I am under a strong belief that amplifying your strengths will always give you a better long term return than fixing your weaknesses.

Truth is, I have lots of quirks. I don't do phone calls (maybe four a year), I don't manage anyone at HubSpot, and I tell dad jokes on stage with 10,000+ people in the audience.

But if I get good enough at what I'm actually strong at, the rest won't matter.

The advice I gave in the post can be summed up by simply finding the thing you're good at, the thing that energizes you. Then double down on it.

Read the full post here.

#5: The evolution of LLMs & what's coming next

This one traced the path from "write me a 500-word article" to "build and launch my startup."

Early LLMs were hilariously bad at simple math. Then came the leap that changed everything: We gave LLMs the ability to call tools. When they encountered a problem requiring calculation, they could hand that task to a Python interpreter and get the correct answer.

Today, if an LLM senses that a prompt could be better solved by coding a simple program, it just does that.

I talked about how we're all becoming "agent managers" instead of traditional knowledge workers. Instead of manually creating spreadsheets, we're learning to communicate objectives to AI systems that execute tasks.

The big prediction: At Anthropic's developer conference, Dario Amodei was asked when we'd see the first billion-dollar company with just one human employee. His response: "2026."

A single entrepreneur, armed with a fleet of AI agents, building and operating a business generating millions in revenue, seems not just possible, but very much likely in the coming months.

Read the post here.

We’re in for a ride in 2026

If I we’re to make a singular prediction (side note, I’ll be posting my 5 predictions for 2026 next week), it’s that agents will get 100x more useful.

We’ll be moving away from hype-y big model breakthroughs and put more focus on AI memory that persists, longer context windows, and most importantly, the infrastructure (the "railways," if you will) for AI agents to actually work together.

There's a lot of understandable debate right now about whether we're in an AI bubble. People are weighing the real economic impact of AI so far and coming up with mixed conclusions.

But if you take a step back and look at the progress from just this year alone, it's been mindblowing. Things are leading up toward the agent-to-agent economy, which in my opinion, is where we’ll see a very measurable difference in the dynamics of AI and work.

We'll be covering it live at simple.ai.

Thank you for being here. Whether you've been reading since the beginning or just joined recently, I'm grateful you're part of this journey. If there are topics you’d like me to cover, just reply or leave a comment. I take requests.

Here's to 2026. Let's keep learning and building.

—Dharmesh (@dharmesh)